Electron spin resonance in tert-butyl-substituted semiquinones. The

Charles Trapp, Charles A. Tyson, and Giovanni Giacometti. J. Am. Chem. ... Edward J. Nanni Jr. , Martin D. Stallings , Donald T. Sawyer. Journal of th...
0 downloads 0 Views 824KB Size
1394

radicals A and B). From the slope of the plot in Figure 4, a difference in the A r r h e n i u s activation energies of 1.8 kcal for these two p r o d u c t s can be determined.

Experimental Section Apparatus and Methods of Analysis. y-Butyrolactone was irradiated in a cycljndrical Cary quartz cell of I-cm path length. The Teflon stopper of the cell was provided with a small hole through which a syringe needle could be inserted for withdrawing samples for gas chromatographic analysis. With this arrangement, a single injection into a chromatograph equipped with a flame detector and a cyanosilicone XF 1500 column (60°, 10 ft X in.) permitted the analysis of allyl formate, cyclopropane, and ethylene. A Carbowax 20M (1 55 ', 10 ft X in.) column was used for the analysis of succinaldehyde. The solutions were all flushed with nitrogen prior to irradiation. This procedure was found to be more than sufficient since the same results were obtained when the solutions were degassed by repeated freeze-thaw cycles or when air saturated. For the analysis of carbon monoxide and carbon dioxide, a different procedure was used. The volatile fraction at -196" was removed with a Toepler pump and its volume determined with a calibrated gas buret. Mass spectrometric analysis showed this fraction to consist entirely of carbon monoxide. A second fraction was taken at -78". This fraction consisted of cyclopropane, carbon dioxide, and ethylene. Mass spectrometric analysis yielded the mole ratio of cyclopropane to carbon dioxide. The cell employed for irradiations in which carbon monoxide and carbon dioxide analyses were performed was larger (3-cm diameter and 0.5-cm path length) and equipped with an approprirate arrangement for degassing.

The experiments were performed with a Hanovia low-pressure mercury lamp (type)eemitting mainly the 2537-A resonance line and with the 2380-A line of a medium-pressure mercury lamp (Hanovia Type A, 500 w). The beam of the resonance lamp was concentrated with a lens and passed through a 3-cm chlorine filter to remove radiation at higher wavelengths, and the light from the medium pressure lamp was rendered monochromatic with a Bausch and Lomb monochromator. For runs above room temperature, the cell was enclosed in an aluminum block furnace. The identification of products was accomplished by the usual instrumental method, and in each case the spectral properties were compared to those of authentic samples. The actinometry employed was ferrioxalate. Materials. The y-butyrolactone was obtained from the Matheson Co. and was fractionated with a spinning-band column, retaining the middle fraction. Examination of the absorption spectrum of the neat lactone (1-cm path length) showed a shoulder beginning at 3100 A. Since this absorption was suspected as being due to impurities, the lactone was subjected to charcoal treatment until all absorption above 2700 A was virtually eliminated. The lactone thus purified shows a sharp rise in absorption beginning at 2600 A. After the charcoal treatments, the lactone was again subjected to a fractionation, the middle fraction being retained. Glpc analysis showed the remaining impurities to be 0.1 %. The cyclohexene was purified by fractionation and chromatography over alumina; cis- and trans-butene (Phillips) were used without further purification. Biacetyl and isopropyl alcohol were purified by fractionation. The acetonitrile used was Matheson "chromato" quality. All solutions were made up immediately after purification and used without delay.

Acknowledgment. This research was supported by Grant AP 00109, National Center for Air Pollution Control. R. S . also wishes to acknowledge support from an Air Pollution Special Fellowship, 1964-1967.

Electron Spin Resonance in t-Butyl-Substituted Semiquinones. The Hyperfine Structure of t-Butyl Protons Charles Trapp, Charles A. Tyson,' and G. Giacometti* Contribution from the Department of Chemistry, Illinois Institute of Technology, Chicago, Illinois 60616. Receioed September 11, 1967

Abstract: The electron spin resonance spectra of three p-benzosemiquinones and four o-benzosemiquinones with t-butyl substituents have been obtained. Resolution of t-butyl proton hyperfine splittings has been accomplished The most reasonable interpretation of the splittings is in terms of carbon-carbon hyperconjugation which produces unpaired spin density in the 6-carbon atom tetrahedral orbitals. Spin density is then transferred t o the protons by a spin-polarization mechanism. The Q value for the latter mechanism is approximately 44 gauss, which indicates that tetrahedral orbitals are more effective than pz orbitals in polarizing hydrogen 1s orbitals. Huckel molecular orbital calculations are performed t o fit the data; the results indicate that the t-butyl group is probably more inductive than the methyl group and that an interaction between the carbonyl group and a t-butyl group ortho t o it may occur.

in all but one of the radicals.

T

he effect o f ring substitution on the electron spin resonance (esr) spectra of s e m i q u i n o n e radical anions has been the subject of a number o f paper^.^-'^ (1) Department of Medical Chemistry, Kyoto University, Sakyo, Kyoto, Japan. (2) National Science Foundation Senior Foreign Scientist Fellow, 1967. University of Padua, Padua, Italy. (3) B. Venkataraman and G. K. Fraenkel, J . Am. Cfiem. Soc., 77,

2707 (1955). (4) R.Hoskins, J . Chem. Pfiys., 23, 1975 (1955). ( 5 ) R. Bersohn, ibid., 24, 1066 (1956).

The effect of c h l o r o and m e t h y l substitution in p - b e n z o semiquinone on the ring proton hyperfine splitting (hfs) can be accounted for by t h e additivity rules o f Venkataraman and Fraenkel.6 The hfs of the m e t h y l protons has been explained in terms of hyperconjugation by B e r ~ o h n . ~Vincow and Fraenkel' did a H u c k e l molecular o r b i t a l (HMO) calculation on p-benzosemiquinones in which the parameters were varied to o b t a i n the best fit for the ring proton hfs. This calculation

(6) B. Venkataraman, B. G. Segal, and G. K. Fraenkel. ibid., 30, 1006 (1959). (7) G.Vincow and G. K. Fraenkel, ibid., 34, 1333 (1961).

JournaI of the American Chemical Society

1 90x5 I March 13, 1968

(8) A. Fairbourn and E. A. C. Lucken, J. Cfiern. Soc., 258 (1963). (9) T.J. Stone and W. A. Waters, ibid., 1488 (1965). (10) L. M.Stock and J. Suzuki, J . Am. Cfiem. Soc., 87, 3909 (1965).

1395

was moderately successful for 2-methyl-p-benzosemiquinones. Fairbourn and Luckens considered the inductive effect of methyl and t-butyl substitution on ring carbon atom spin densities in p-benzosemiquinone using HMO theory. Good agreement of the calculated values with experimental results was not obtained, but their treatment did indicate that the inductive effect of the t-butyl group was slightly greater than that of methyl and that an interaction of the substituent with the carbonyl group might account for some of the observed discrepancies. This “ortho” interaction effect would essentially result in a shortening of the carbon-oxygen bond distance. Stone and Watersg noted that very few substituted o-semiquinones had been studied by esr. They report results on a large number of compounds, some of which were methyl- and t-butyl-substituted o-semiquinones. They did not, however, attempt a theoretical interpretation of the spectra. Stock and Suzuki’O have done esr on 2-t-b~ty1-C’~-and 2-trifluoromethyl-p-benzosemiquinone and have shown that spin density is transferred to /3 nuclei presumably by way of a hyperconjugative mechanism. They observed @-Cl 3 hyperfine splittings in enriched samples and were able to estimate the spin density in 2s orbitals of the @-carbonatoms. They concluded that carbon-carbon hyperconjugation is only slightly less effective than carbon-hydrogen hyperconjugation. Hoskins4 has observed an esr spectrum of 4-t-butylo-benzosemiquinone. The large number of lines in the spectrum indicated the presence of t-butyl proton splittings, but because of limited resolution the splitting constant could not be obtained. Fraenkel” has reported that it is possible to resolve the proton hfs in 2,5-di-t-butyl-p-benzosemiquinone.In view of this we conclude that it may be possible to observe t-butyl splittings in a larger number of t-butyl-substituted pand o-benzosemiquinones. In this paper we report the esr spectra of seven t-butylsubstituted benzosemiquinones. The t-butyl proton hfs has significance from the point of view of carboncarbon hyperconjugation. At present very little good experimental data are available which are directly associated with this effect. Any t-butyl proton hfs in one of the radicals, 2-t-butyl-p-benzosemiquinone, may be related to the @-C13hfs observed in the same radical by Stock and Suzuki. lo The mechanism of the transfer of spin density from the @-carbonatoms to the t-butyl protons is probably through a spin-polarization mechanism of the type which is effective for ring protons. We will attempt to determine a Q value for this interaction from our experimental data in conjunction with the data of Stock and Suzuki on @-C13hfs splittings and also from a Huckel molecular orbital model. The presence of an interaction between the t-butyl group and the carbonyl group, when they are ortho to each other, should probably affect the nature of the t-butyl hfs, and we should find some evidence for this in the spectra and in the HMO parameters used to fit the hfs data. In addition we shall attempt to determine whether the additivity relationships of Venkataraman and Fraenkel,3 which apply to methyl-substituted benzosemiquinones, will work in the case of t-butyl substitution. Finally, we shall attempt to fit (11) G. K. Fraenkel, Ann. N. Y. Acad. Sci., 67, 546 (1957).

the experimental data on both methyl and t-butylsubstituted o- and p-benzosemiquinones using HMO theory and a conjugative model for the substituent, Experimental Section The esr spectra were all taken on a Varian Associates V-4502-12 epr spectrometer. All measurements were made at 25” in alkaline methanol-water solutions. The methanol concentration was generally 50% by weight, but in some cases lower methanol concentrations were necessary to get full resolution. The splitting constants changed only slightly with change in methanol concentration. The p- and o-semiquinones were generated from the corresponding hydroquinone or catechol in a Pyrex cell by bubbling air through the alkaline solutions. The reaction mixture was circulated through 1.5-mm i.d. Pyrex tubing in the microwave cavity using a peristaltic pump. Flow rate was 30 ml/min, and the pH reading was kept constant at about 10.0-10.5, relative to standardization in -log [H+] units in 50% methanol.12 Concentrations of the radicals were estimated at about M. Chemicals were all commercially available: 2-t-butylhydroquine from Eastman Organic Chemicals; 2,5-di-t-butylhydroquinone, 4-t-butylpyrocatecho1, 3,5-butyl-5-methylpyrocatechol, and 5-2butyl-3-methylpyrocatechol from Aldrich Chemical Co. ; 2,6-dit-butyl-p-benzoquinone, a gift from the Ethyl Corp., and 3-5-di-1butylpyrocatechol from Gallard-Schlesinger, Long Island City, N. Y . All were used without further purification except for the last item which was recrystallized twice from benzene. The hydroquinone of 2,6-di-t-butyl-p-benzoquinone was prepared by zinc-HC1 reduction of thep-quinone.I3 A frequency meter and an nmr gaussmeter were used to measure g values of the radicals studied. It was not felt that the g values of these radicals were significant so that they were measured to an accuracy of only one part in 3000. The g values of all the radicals were constant at 2.0042 i 0.0007. The HMO calculations were performed on an IBM Model 360-40 data-processing system with a Jacobi diagonalization routine.

Results 2-t-Butyl-p-benzosemiquinone. Our results are very close to those reported by Stock and Suzuki’O for the ring protons. The splitting constants are given in Table I. The subscripts on the splitting constants refer to the ring carbon atom to which the hydrogen atom: is attached. The coupling constants are assigned on the basis of the results of the HMO calculations described in the next section and are, therefore, not unambiguous. u3 and a6 tend to decrease about 10% from the values given in the table when the methanol concentration is reduced to 12.5 %, whereas us is unaffected. The spectrum consists of eight main lines as previously reported. l o Upon higher resolution each main line is split into a number of lines. This latter splitting is produced by the f-butyl protons, and the splitting constant is uZf-Bu= 0.060 f 0.005 gauss. The three main lines on the low-field side of the center of the spectrum are shown in Figure 1. The presence of a center line in the group associated with each main line may indicate that the t-butyl group is not freely rotating. This lack of free rotation may be evidence of an interaction between the t-butyl group and the carbonyl group ortho to it. 2,5-Di-t-butyl-p-benzosemiquinone.Fraenkelll reports the observation of t-butyl proton hfs in this radical, and Adams, Blois, and Sands14 determined the ring proton hfs splittings. Our experimental values are given in Table I and agree with those previously reported. The spectrum is shown in Figure 2 and is a (12) C. A. Tyson and A. E.Martell, J. Am. Chem. SOC.,in press. (13) A. I. Vogel, “A Textbook of Practical Organic Chemistry,” 3rd ed, John Wiley and Sons,Inc., New York, N . Y., 1956, p 748. (14) M. Adams, M. S. Blois, and R. H. Sands, J . Chem. Phys., 28, 774 (1958).

Trapp, Tyson, Giacometti

Esr in t-Butyl-Substituted Semiquinones

1396 Table I. Hfs Splitting Constants of r-Butyl- and Methyl-Substituted Benzosemiquinonesa Radical

Solventb

2-t-Butyl-p2-t-Butyl-p2,5-Di-r-butyl-p2,6-Di-r-butyl-p4-r-Butyl-03,5-Di-r-butyl-o3-Methyl-5-t-butyl-o3-r-Butyl-5-methyl-o-

50% CHIOH 12.5%CH3OH 50ZCHsOH 50% CHaOH 50% CHIOH 50% CHaOH 50% CHIOH 50% CHIOH

2-Methyl-p-d 2-t-Butyl-p-e 2,5-Dimethyl-p-d 2,6-Dirneth~l-p-~ 2,5-Di-t-methyl-p-/ 4-Methyl-o-g 4-r-B~tyl-o-~ 3,5-Di-r-butyl-o-i 3-r-Butyl-5-methyl-o-~

7 0 % CHICHZOH

Acetonitrile 70% CHICHZOH 7 0 z CHICHZOH Ethanol-water Water Ethanol-water Acetone Water

u3

a2

A. This Work 0.060 i 0.005 1.55 f 0.02 0.060 f 0.005 1.41 i 0.02 0.060 f 0.005 2.07 f 0.02 0.055 f 0.005 1.17 f 0.02 ... 0.3 f 0.1